What happens when you put a loudspeaker in a corner?

[maxro]

That's a good demonstration. And, by the thread title, I was expecting a lame question by some noob like me.

However, that speaker is not actually in the corner, it is near the corner.

What happens when you back the speaker right into the corner, so that its front is at 45 degrees to the side walls? I figure a front-loaded horn effect to direct and amplify the sound wavefront. It would at the very least get rid of baffle step. I suppose this would be acoustically superior to the mirror picture.(?)

Max

 

[Romberry]

Actually, a corner can be a very good place

Roy Allison of Allison Acoutics designed and built his full range Model Three explicity for corner placement. In a corner, the Allison Three was the eqyuivalent of the Allison One which was designed to go against the wall at the intersection of the floor and wall. If you do a bit of searching on the net and through the archives at the JAES, you'll find Allison's research on room/speaker interaction and how to best work with/take advantage of the factors that often lead to peaks and dips in the radiated power through the bass region. Anyway, a corner *can* be a great place for a speaker that's designed to go there.

 

[lndm]

What happens when you back the speaker right into the corner, so that its front is at 45 degrees to the side walls? I figure a front-loaded horn effect to direct and amplify the sound wavefront. It would at the very least get rid of baffle step.

That's a good question and I agree with your assumtions. What you might do now is model a conical horn, to see whether you like it.

Other options could be making a flat baffle to support the sound down to a reasonably low frequency, and fitting it into the corner like you suggested. Even better, you could build a horn that fits into the corner so the corner becomes a part of it.

You would certainly get rid of the baffle step and enjoy increased sensitivity (and more).

I for one don't subscribe to the theory that a baffle must be narrow for good imaging. It really only changes the frequency of the inevitable. Why not take it down all the way?

 

[Svante]

This is correct in theory, so we can say that the corner makes one driver in free space with 'x' SPL sound like there are eight of them, and since for every doubling of identical SPL there's a ~3.01 dB increase, and as Poobah noted, this occurs 3x, we wind up with a theoretical ~ +9.03 dB.

GM

Now let's get a few things straight. 🙂

Sound pressure p is measured i Pascal (Pa).
Sound pressure level (SPL) Lp is measured in decibel (dB).

The sound pressure level can be calculated knowing that the reference for SPL is pref=20 µPa, as:

Lp=20*log10(p/pref)

That means that e.g. 1 Pa corresponds to 94 dB.

To think of doubling the SPL is rarely meaningful, eg going from 50 to 100 dB.

However, doubling the sound pressure (in Pa) is exactly what happens with one reflective surface and long wavelengths.

Doubling the sound pressure leads to 20*log10(2) = 6 dB gain.

So, for low frequencies:
one plane yields 6 dB gain
two planes yield 12 dB gain
three planes yield 18 dB gain

Otoh, if the frequency is high, the sound from the sources will not occur perfectly in phase. If the frequency is high enough, and if one looks at an average over many directions, one can look at the intensity, rather than the pressure.

Sound intensity J is measured in W/m²,
Sound intensity level (SIL) Lj is measured in dB

The sound intensity level can be calculated knowing that the reference for SIL is Jref=1pW, as:

Lp=10*log10(J/Jref)

Doubling the sound intensity leads to 10*log10(2) = 3 dB gain.

So, for low frequencies:
one plane yields 3 dB gain
two planes yield 6 dB gain
three planes yield 9 dB gain

The important thing here is to notice the difference between high and low frequencies.

What I wrote above can be shown graphically as in this image. d is the source to wall distance. Apologies for the swedish text.

An externally hosted image should be here but it was not working when we last tested it.